Side profile facial recognition using CNN

Abstract

Facial recognition is rapidly advancing in the field of Artificial Intelligence (AI) and according to Statista, the market size is projected to reach US$5.71B in 2024 for facial recognition [1]. It is used across various domains such as security, law enforcement, and healthcare. The current systems work well with the front view of a person as only the frontal image is used in the training data of these systems. However, it is challenging in real-world scenarios as it is common for people to be seen from various angles. Thus, a robust facial recognition system that could identify a person from side and front would reduce challenges faced in various applications. This project researched and trained a side profile facial recognition system using CNN to learn the unique facial features.